paperKB
coga / coga-kb
Help
Sign in

Chunk #69 — Methods — Runtime analysis

Source
scCODA is a Bayesian model for compositional single-cell data analysis.
Embedded
yes

Text

For five cell types, datasets of all tested sample sizes require about 0.0025 s per HMC iteration on average. The time per iteration increased linearly with the number of cell types for all sample sizes. This effect is more pronounced for larger sample sizes, with 40 total samples (20 per group) and 50 cell types requiring the longest average time per step of about 0.0035 s, while the average runtime per step for datasets with five samples was always below 0.0027 s. Thus, running scCODA with the default number of 20,000 HMC iterations on any dataset of typical size should produce results within a few minutes.